Subject1: An Order-Invariant Sparse Inverse Covariance Matrix Estimation Based on the Modified Cholesky Decomposition
Lecturer: Kang Xiaoning
Contents: The modified Cholesky decomposition is commonly used for inverse covariance matrix estimation given a specified order of random variables. However, the order of variables often is not available or cannot be pre-determined. In this work, we propose a novel order-invariant estimator for high-dimensional sparse inverse covariance matrix based on the modified Cholesky decomposition. The proposed method efficiently ensembles a set of estimates obtained from multiple orders of random variables, and by using thresholding technique appropriately it encourages the sparse structure in the estimate. The proposed method not only provides an accurate estimation, but also can effectively capture the underlying structure of the inverse covariance matrix. The consistent property is constructed under some weak regularity conditions. Simulation studies show the superior performance of the proposed method in comparison with other approaches. We also apply the proposed method into the linear discriminant analysis for analyzing classification examples.
Starting Time:4:00-5:00pm May 26, 2017
Location: The conference room on the second floor of Technology lab building
About Prof. Yu Sui：Large sparse inverse covariance matrix estimation and its application、High-dimensional data analysis and modeling、Bayesian hierarchical modeling、Semiparametric modeling and inference。